The production of magnesium metal from its oxide by carbothermal reduction has been well-known for nearly a century. Fundamental to the process is the rapid quenching of the reaction products (carbon monoxide and magnesium vapour) to a temperature below that at which the reversion reaction takes place (about 400° C.). One way that has been proposed to achieve the requisite quenching has been to eject the product gases through a convergent-divergent nozzle at supersonic speed. This results in rapid expansion of the gases and instantaneous cooling as required (estimated to be at a rate of up to 105° C.s−1). Examples of this approach include the disclosures of Hori (U.S. Pat. No. 4,147,534 and U.S. Pat. No. 4,200,264). In order to avoid the reversion reaction Hori teaches that thermal control of the product gases is important throughout the process from the reaction chamber to the product collection point via the nozzle.
Notwithstanding the general approach recommended by Hori, it has been found that solids tend to be deposited and accumulate on internal surfaces of the nozzle that are in contact with the gaseous products flowing through the nozzle. This can lead to deterioration in the performance of the nozzle and, even worse, blocking. Blocking results in potentially unsafe operating conditions (due to creation of over-pressure) and it then becomes necessary to shut down production and re-bore or replace the nozzle. The disclosures of Hori do not report blocking, and the reasons for this are not entirely clear. It may be because the processes were operated only on a small scale and/or with relatively pure reactants (impurities can add to the blocking problem). Interestingly, a methodology such as that proposed by Hori has not been implemented on a commercial scale.
It is also relevant to mention the disclosure by Donaldson, A and Cordes, R A, Rapid Plasma Quenching for the Production of Ultrafine Metal and Ceramic Powders, JOM, 2005:57(4), pp. 58-63. This describes pre-heating of a quench nozzle in the experimental production of aluminium from a plasma reactor. Pre-heating takes place on start-up by feeding hot argon from the reaction furnace through the nozzle. Under sonic conditions, such heating would at best result in the nozzle surfaces reaching a temperature at equilibrium with the gas stream travelling through it. The present inventors have found that pre-heating the nozzle using a gas stream, as per Donaldson and Cordes, may be insufficient to produce and maintain the required temperature to avoid deposition problems, especially in the production of metals other than aluminum, for instance magnesium.
Further, Donaldson and Cordes mention pre-heating of the nozzle on start up only. Presumably, thereafter the temperature of the exit gas from the reaction furnace is relied upon to maintain nozzle temperature. However, the present inventors have found that this is not a reliable way to proceed to avoid deposition problems.
Against this background, it would be desirable to provide a process and reactor that enables the carbothermal reduction process described to be implemented on a commercial scale and that enables deposition problems to be alleviated and preferably avoided altogether for production of a range of metals, especially magnesium.